The structure of a distributed network comprises a multiplicity of system nodes which are interconnected via a network structure, for example in a client-server relationship, and provide functionalities which cannot be provided by means of an individual component. Operator stations, engineering stations, archiving stations, control devices with processing stations, input and output units and communications interfaces, for example, are used as system nodes in the distributed network.
The functional components are normally distributed among a multiplicity of system nodes, also referred to below as a computer or data processing device, which are interconnected via one or more physical networks. Operating systems, application programs, control programs and various functional components of the system are installed on the system nodes, whereby the individual nodes with their various functional components are dependent on one another in different ways.
In order to upgrade the system software, which is stored in software packages, a new or improved system version is installed on the computers of the distributed system, while retaining the user-specific applications, such as, for example, project data and settings, whereby the functional components of the existing system are replaced with new components and the upgraded system is able to provide the same, albeit modernized, fundamental functionality. This procedure is referred to below as a system upgrade.
Methods currently employed use, for example, installation programs, referred to as software installers, which perform the system upgrade on a single computer of the distributed system. Prior to the installation of the new programs, these installation programs check the existing software configuration on the computer and, in the event of a re-installation, suggest alternatives according to the available upgrade facilities of the software.
A detailed knowledge of the dependencies between the individual system nodes is required in order to take into account the dependencies between the individual system nodes of the distributed automatic system when the software packages are upgraded.
Installation manuals often offer only a general knowledge of the existing dependencies between the distributed system software. The planning of the system upgrade and the identification of the existing system components which need to be replaced with upgraded new components must be carried out manually by the user.
Comprehensive expert knowledge is required to install the new software packages on the various system nodes and configure the relationships and dependencies between the system nodes, since the procedure for installing and configuring the software packages is often only inadequately described and is difficult to follow in the guidelines and instructions prescribed for that purpose.
Even existing analysis tools consider only compatibilities of individual software packages and their versions with one another on one computer, but not logical system-related dependencies, such as, for example, in relation to computer-independent software packages.
No dependency rules can be formulated for planning an automatic upgrade for all system nodes of the distributed system.
Newer versions of the complete system software are normally offered in a plurality of software packages, which the user himself must identify in order to ascertain which system components need to be upgraded and installed on which nodes, or which system components need to be replaced with a new version.
The system structure plans and the identification of the system components for a complete and correct installation of the interrelationships and interdependencies of the system components are also manually created and checked separately for each system component, whereby the administrative effort involved in upgrading the software packages is highly time-consuming and increases in proportion to the scope of the system topology.
An automatic installation, configuration and system diagnosis of the distributed system with an upgraded, newer version of the system software of the distributed automatic system over an existing version cannot be carried out with the methods currently employed.